OBJECTIVE: The study was undertaken to identify early pregnancy vaginal markers predictive of subsequent preterm birth. STUDY DESIGN: In a multicenter Bacterial Vaginosis (BV) Trial, 21,554 women were screened with a vaginal pH and of these, two populations were studied. These included 12,041 who had a pregnancy outcome in the database and 6838 women who had a vaginal pH of 4.5 or greater and a Gram stain score and a pregnancy outcome in the database. ColorpHast Indicator Strips were used to determine the vaginal pH and the Nugent criteria were used to determine a vaginal Gram stain score of 0 to 10. RESULTS: Delivery at <37, <35, or <32 weeks' gestation was similar for women with a vaginal pH of less than 4.4 or 4,7 (P not significant) but was increased in women with a pH of 5.0 (P = .04, .02, .03, respectively) or with a pH of 5.0 or greater (at each gestational age P < .0001). The effect of pH of 5.0 or greater was similar for women who had a spontaneous preterm birth at each gestational age (P < .0001) or birth weight of less than 2500 g or less than 1500 g (P < .0005). Women with a vaginal pH of 4.5 or greater and a Gram stain score of 9 to 10 (compared with 0-8) had increased preterm births at <37, <35, and <32 weeks' gestation (P < .01), and birth weights less than 2500 g (P < .0001) or less than 1500 g (P < .01). Women whose vaginal pH was 5.0 or greater had a higher prevalence of vaginal fetal fibronectin ≥50 ng/mL (P < .0001), but the proportion of women with a vaginal fetal fibronectin ≥50 mg/mL did not differ by Gram stain score. CONCLUSION: Women with a vaginal pH of 5.0 or greater or a vaginal pH of 4.5 or greater and a Gram stain score of 9 to 10 had significantly increased preterm births at <37, <35, and 32 weeks' gestation and/or a birth weight less than 2500 g or less than 1500 g.